Eli Collins, vice president of product management at Google DeepMind, began rolling out AI-generated video tools for the company’s board of directors back in 2022. Despite the slow speed of the model, the cheap cost of operation, and sometimes going off-kilter, he says it was an eye-opening moment for them to see new video clips made from random information.
Now, a few years later, Google has announced plans for a tool within the YouTube app that will allow anyone to create AI video clips, using the company’s Veo model, and post them directly as part of YouTube Shorts. “Looking forward to 2025, we’ll allow users to create standalone video clips and shorts,” said Sarah Ali, YouTube’s senior director of product management. “They will be able to produce six-second videos with an open message.” Ali says that this update can help creators who are hunting for images to fill a video or try to see something amazing. He maintains that the Veo AI tool is not intended to replace innovation, but to augment it.
This isn’t the first time Google has introduced production tools for YouTube, though this announcement will be the company’s most extensive AI video integration to date. In the summer, Google launched a test tool, called Dream Screen, to generate an AI background for videos. Ahead of next year’s full rollout of manufactured clips, Google will update that AI green screen tool with the Veo model sometime in the next few months.
The leading technology company has shown many AI video models in recent years, such as Imagen and Lumiere, but it is trying to combine a more unified vision with the Veo model. “Veo will be our model, in a way, going forward,” Collins said. “You shouldn’t expect five more models from us.” Yes, Google may release another video model eventually, but expect to focus on the Veo in the near future.
Google is facing competition from many startups building their own text-to-video production tools. OpenAI’s Sora is a well-known competitor, but the AI video model, which was announced in early 2024, is not yet publicly available and is limited to a small number of testers. In terms of widely available tools, AI startup Runway has released several versions of its video software, including the latest tool to edit original videos into original versions of the clip.
YouTube’s announcement comes as AI-powered productivity tools have grown increasingly contested by creators, who sometimes view the current wave of AI as stealing from their work and trying to undermine the creative process. Ali doesn’t see productive AI tools coming between creators and the authenticity of their relationship with viewers. “It’s about the audience and what they’re interested in—not necessarily about the tools,” he says. “But, if your audience is interested in how you did it, that will open up with meaning.” Google plans to tag all AI-generated YouTube Shorts videos with SynthID, which embeds an invisible tag to help identify a video as synthetic, and includes the disclaimer “made with AI” in the description.
Hustle-culture promoters are already trying to game the algorithm by using many third-party tools to automate the creative and monetization process with minimal effort. Will next year’s Veo integration lead to a new outbreak of low-quality, spammy YouTube shorts dominating users’ feeds? “I think our knowledge of recommending the right content to the right viewer is applicable in this AI-level world, because we’ve been doing it on such a large scale,” Ali said. He also points out that YouTube’s general guidelines still apply regardless of the tool used to create the video.
AI creativity often has a unique aesthetic, which may be relevant to video creators who value individuality and want their content to feel unique. Collins hopes Google’s thumbprints aren’t all over the place for video AI. “I don’t want people to look at this and say, ‘Oh, that’s the DeepMind model,'” he says. Getting AI output data to match what the creator envisioned is the primary goal, and getting away from Veo’s obvious beauty is essential to achieving wider flexibility.
“A big part of the journey is actually building something that’s useful to people, that’s scalable, and that’s easy to use,” Collins said. “It’s not just a demo. It is used in the actual product.” He believes that putting AI productivity tools right into the YouTube app will change creators, as well as DeepMind. “We have never made a product for creators,” he said. “And we’ve certainly never done it at this level.”